NETWORK CONTROL SYSTEM

Description

TECHNICAL FIELD

[0001] The present invention relates to a network control system for packet-transmitting a manipulated variable (referred to as "MV" hereinafter) of a controller, which is calculated based on a deviation between a processed variable (referred to as "PV" hereinafter) from a sensor that measures a physical quantity of a controlled object plant and a set value (abbreviated to "SV" hereinafter) at a predetermined sampling period, to an actuator provided on the controlled object plant side via a network, and causing the actuator to provide the MV to the controlled object plant.

BACKGROUND ART

[0002] The trend of plant control is directed toward the networked control using a wireless communication or the like in the future. In the networked control, a packet transmission error that the PV or the MV does not reach a destination from a sender due to aggravation of communication circumstances, i.e., a packet loss, occurs. In the network control system, a layout is demanded in view of particularly a packet loss.

[0003] FIG.8 is a functional block diagram showing a configurative example of the network control system in the prior art. As a basic configuration, a plant side 10 and a controller side 20 are connected mutually via a network 30.

[0004]  The PV is transmitted from a sensor 12, which measures a physical quantity of a controlled object plant 11 on the plant side 10, to a controller 21 on the controller side 20 via the network 30. The controller 21 calculates a deviation between the transmitted PV and the set SV, e.g., a PID-operated MV, and transmits this MV to an actuator 13 on the plant side 10 via the network 30. The actuator 13 provides the received MV to the controlled object plant 11.

[0005] When the controller 21 cannot receive the PV due to the packet loss, the controller 21 compensates for the PV required for the MV calculation by a PV complementary value provided by a PV complementing unit 22. Similarly, when the actuator 13 cannot receive the MV due to the packet loss, the actuator 13 compensates for the MV, which is to be provided to the controlled object plant 11, by an MV complementary value provided by an MV complementing unit 14.

[0006] As the PV complementary value and the MV complementary value, the latest PV and MV values received in the sampling period immediately before the packet loss occurs or the estimated values calculated based on trend data in predetermined sampling periods are employed.

[0007] 

Patent Document 1: JP-A-08-328650

Patent Document 2: JP-A-11-345003

Patent Document 3: JP-A-2002-040917

Patent Document 4: JP-A-2006-277306

Patent Document 5: JP-A-2007-233891

DISCLOSURE OF THE INVENTION

PROBLEMS THAT THE INVENTION IS TO SOLVE

[0008] In the network control system having the conventional configuration, the trouble caused due to an occurrence of the packet loss that the PV or the MV does not reach the destination from the sender depending on the communication circumstances is not fully taken into account. Therefore, the serious problems described hereunder arise in the plant control at a time of occurrence of the packet loss.

[0009] In the case where either the MV that the controller 21 calculates on account of the packet loss and the MV that is provided actually to the controlled object plant 11 are different or the actual PV and the PV that is complemented by the PV complementing unit are different, an internal state of the controller deteriorates and thus the response deteriorates.

[0010] FIG.9 is a characteristic view showing deterioration in a response characteristic when a packet loss occurs. The MV that the controller 21 calculates when the packet loss occurs and the MV that is provided actually into the controlled object plant 11 are different, and a response is deteriorated. In this example, the responses are given when the SV is changed from 20 to 40. Here, a solid line indicates the characteristic when no loss occurs, and a dot-dash line indicates the characteristic when the actuator 13 cannot receive the MV due to the packet loss at a time of changing the set value.

[0011] When the SV is changed, the controller 21 calculates the MV in response to the change. In this event, since the actuator 13 cannot receive the MV that is calculated in the controller 21 upon changing the set value, a start of the response of PV is delayed in the response indicated with a dot-dash line.

[0012] In this case, the controller 21 increases an integrated value indicating its own internal state during a time in which a start of the response of PV is delayed. As a result, the integrated value is excessively largely increased. This excessively largely integrated value must be restored to a proper value.

[0013] When the integrated value becomes larger than an appropriate set value, the integrated value cannot be restored to the proper value unless the PV is increased larger than the SV once. Therefore, an overshoot of the PV is increased. With the above, a disturbance of the response is generated.

[0014] The present invention has been made to solve the problems, and it is an object of the present invention to implement a network control system capable of realizing stable plant control by compensating a controller in view of an occurrence of a packet loss.

[0015] This object is achieved by the features as set forth in claim 1. Further advantageous embodiments of the present invention are set forth in the dependent claims.

(1) A network control system for packet-transmitting a manipulated variable of a controller, which is calculated based on a deviation between a process variable from a sensor that measures a physical quantity of a controlled object plant and a set value at a predetermined sampling period, to an actuator provided on the control object plant side via a network, and causing the actuator to provide the manipulated variable to the controlled object plant, the network control system comprising:

an MV complementing unit that provides a complementary value of the manipulated variable to the actuator when an error occurs in a packet transmission;

an MV buffering unit that answers back trend data of the manipulated variable being provided to the controlled object plant from the actuator or the complementary value being complemented by the MV compensating unit for the actuator, to the controller via the network; and

an MV compensating unit that corrects a calculation of the manipulated variable of the controller based on the trend data or the complementary value being answered back.

(2) The network control system according to item (1), wherein the MV compensating unit calculates a virtual SV value based on the trend data or the complementary value being answered back and the trend data being held on the controller side, such that an amount of manipulated variable obtained from a sampling at a time of occurrence of an error in the packet transmission to a current sampling coincides with a value of the trend data, and then corrects/calculates the manipulated variable in the controller from the virtual SV value.

(3) The network control system according to item (1), wherein the controller stops temporarily the transmission of the manipulated variable to the actuator until the controller acquires an answer-back of the trend data or the complementary value.

(4) The network control system according to item (1), wherein the controller transmits separately a first manipulated variable corresponding to an amount of change in the set value in a differentiating operation and a second manipulated variable corresponding to an amount of change in the set value except the amount of change in the set value in the differentiating operation, and the MV complementing unit provides a sum of the first manipulated variable and the second manipulated variable to the actuator as the complementary value when the packet transmission is in a normal state, and provides only the second manipulated variable to the actuator as the complementary value when the packet transmission is in an error state.

(5) The network control system according to item (1), wherein when the processed variable measured by the sensor is packet-transmitted to the controller via the network, the network control system further comprises:

a PV complementing unit that provides a complementary value of the processed variable to the controller when an error occurs in the packet transmission; and

a PV buffering unit that transmits trend data of the processed variable measured by the sensor to the controller via the network.

(6) The network control system according to item (1), wherein the PV complementing unit corrects the complementary value of the processes variable based on the trend data of the manipulated variable.

ADVANTAGES OF THE INVENTION

[0016] According to the configuration of the present invention, even though such a situation is caused under network control that the actuator cannot receive the MV due to the packet loss, the stable plant control can be carried out by making the MV compensation on the controller side not to largely disturb the plant state.

BRIEF DESCRIPTION OF THE DRAWINGS

[0017] 

FIG.1 is a functional block diagram showing an embodiment of a network control system according to the present invention.

FIG.2 is a flowchart showing procedures for compensating a calculation of a manipulated variable of a controller.

FIG.3 is a flowchart showing other procedures for compensating a calculation of a manipulated variable of a controller.

FIG.4 is a flowchart showing procedures for setting an MV complementary value according to the present invention.

FIG.5 is a functional block diagram showing another embodiment of a network control system according to the present invention.

FIG.6 is a functional block diagram showing still another embodiment of a network control system according to the present invention.

FIG.7 is a functional block diagram showing yet another embodiment of a network control system according to the present invention.

FIG.8 is a functional block diagram showing yet a configurative example of a network control system in the prior art.

FIG.9 is a characteristic view showing deterioration in a response characteristic when a packet loss occurs.

FIG.10 is a flowchart showing procedures for setting an MV complementary value in the prior art.

DESCRIPTION OF REFERENCE NUMERALS AND SIGNS

[0018] 

100 plant side

101 controlled object plant

102 sensor

102a PV buffering unit

103 actuator

103a MV buffering unit

104 MV complementing unit

200 controller side

201 controller

201a MV buffering unit

201 b PV buffering unit

201c MV compensating unit

202 PV complementing unit

202a process model

300 network

BEST MODE FOR CARRYING OUT THE INVENTION

[0019] The present invention will be explained with reference to the drawings hereinafter. FIG.1 is a functional block diagram showing an embodiment of a network control system according to the present invention. A basic configuration is identical to that of the conventional system explained in FIG.7.

[0020] The PV is transmitted from a sensor 102, which measures a physical quantity of a controlled object plant 101 on a plant side 100, to a controller 201 on a controller side 200 via a network 300. The controller 201 calculates a deviation between the transmitted PV and the set SV, e.g., PID-operated MV, and transmits this MV to an actuator 103 on the plant side 100 via the network 300. The actuator 103 provides the received MV to the controlled object plant 101.

[0021] When the controller 201 cannot receive the PV due to the packet loss, the controller 201 compensates for the PV required for the MV calculation by the PV complementary value provided by a PV complementing unit 202. Similarly, when the actuator 103 cannot receive the MV due to the packet loss, the actuator 103 compensates for the MV, which is to be provided to the controlled object plant 101, by the MV complementary value provided by an MV complementing unit 104.

[0022] An MV buffering unit 201 a is provided in the controller 201. This MV buffering unit 201a holds trend data having (m+1) pieces of MV information that the controller 201 calculated in respective samplings from the current sampling to the m samplings ahead of this current sampling.

[0023] Similarly, an MV buffering unit 103a is provided in the actuator 103. This MV buffering unit 103a holds trend data having (m+1) pieces of MV information that the actuator 103 provides actually to the controlled object plant 101 in respective samplings from the current sampling to the m samplings prior to this current sampling.

[0024]  The actuator 103 transmits the trend data having (m+1) pieces of information being held in the MV buffering unit 103a to the controller 201 via the network 300 to perform the answerback.

[0025] When the actuator 103 cannot receive the MV being transmitted from the controller 201 due to the packet loss, the actuator 103 compensates for the MV by the MV complementary value provided by the MV complementing unit 104. At this time, the MV being provided actually to the controlled object plant 101 is different from the MV that the controller 201 calculated.

[0026] This divergence of MV disturbs the internal state of the controller 201, and act as a factor to worsen the subsequent responses. Therefore, the controller 201 compares MV trend data being received by an answerback from the actuator 103 with MV trend data that MV buffering unit 201 a itself holds therein. When both trend data are different, the compensating unit 201 c performs the compensating operation by utilizing the MV trend data answered-back, and corrects the own internal state of the controller 201.

[0027] A PV buffering unit 102a is provided in the sensor 102. This PV buffering unit 102a holds trend data having (m+1) pieces of PV information that the sensor 102 measured in respective samplings from the current sampling to the m samplings prior to this current sampling.

[0028] Similarly, a PV buffering unit 201b is provided in the controller 201. This PV buffering unit 201b holds trend data having (m+1) pieces of PV information that the controller 201 employs in the MV calculation in respective samplings from the current sampling to the m samplings prior to this current sampling.

[0029] The sensor 102 transmits the trend data having (m+1) pieces of information held in the PV buffering unit 102a to the controller 201 via the network 300.

[0030] When the controller 201 cannot receive the PV being transmitted from the sensor 102 due to the packet loss, the controller 201 compensates for the PV by the PV complementary value provided by the PV complementing unit 202. In this case, this completed value is different from the true PV.

[0031] Next, an embodiment of a compensating operation of the MV compensating unit 201c will be explained hereunder. A manipulated variable u_n (= MV) obtained when the controller 201 performs the PID operation on a deviation between PV and SV is given by Eq.(1).

where e_n : deviation (= r_n - x_n)

r_n : SV

x_n : PV

Δt : sampling time

K_p : control gain

T_l : integration time

T_n : differentiation time

[0032] The MV compensating unit 201c calculates a virtual SV value based on the trend data being answered back and the trend data being held on the controller 201 side such that an amount of manipulated variable obtained from the sampling at a time of occurrence of an error in the packet transmission to the current sampling coincides with the value of the trend data being answered back. Then, the MV compensating unit 201 c corrects/calculates the MV in the controller 201 from this virtual SV value. This virtual SV value is given by Eq.(2).

[0033] FIG.2 is a flowchart showing procedures for compensating a calculation of the manipulated variable of the controller. In step S1, it is found that the MV calculated by the controller is different from the MV that is provided actually to the plant. Then, in step S2, it is checked how many samplings prior to the current sampling at a maximum the MV that the controller holds therein (the MV calculated by the controller or the MV updated by the compensating operation) firstly becomes different from the MV that is provided actually into the plant, and this sampling number is assumed as k.

[0034] Assuming that the current sampling is n, the compensating operation is started from the (n-k)-th sampling, in which the MV that the controller holds therein becomes different at first from the MV that is provided actually into the plant.

[0035] In step S3, the MV, which is provided actually into the plant in the (n-k)-th sampling as the compensated object sampling, is substituted into u_n-k shown in Eq.(2), and thus a set value r_n-k appropriate to the MV is calculated.

[0036] In step S4, the calculated r_n-k is considered as the set value of the (n-k)-th sampling, and then a deviation in the (n-k)-th sampling and an integrated value up to the (n-k)-th sampling are calculated once again.

[0037] Then, in step S5, the similar operation is performed by applying k = k-1, i.e., by shifting the compensated object sampling to the next sampling. Then, in step S6, the similar operation is repeatedly performed until the deviation in the (n-k)-th sampling and the integrated value up to the (n-k)-th sampling are calculated (k = 0).

[0038] In step S7, the MV that the controller holds therein is updated to the MV that is used for the compensation and is provided actually into the plant. In step S8, the compensating process is ended.

[0039] In FIG.2, the controller 201 is directed to a position-type PID controller. The controller 201 is not limited to the position-type PID controller, and the controller 201 is also directed to a speed-type PID controller. In the speed-type PID controller, the controller 201 transmits a difference Δu_n of the MV shown in Eq.(3) to the actuator 103.

[0040] When this difference Δu_n is lost due to the packet loss, the actuator 103 makes up any Δu_n. For example, when the MV in the preceding sampling is maintained as it is, Δu_n = 0.

[0041] Then, the difference applied to the MV being provided actually to the plant becomes different from the difference calculated by the controller, and thus the internal state of the controller gets out of order, likewise the position-type PID controller. Therefore, the compensation is performed similarly to the case of the position-type. A compensating equation is given by Eq.(4).

[0042] FIG.3 is a flowchart showing procedures for compensating the MV calculation in the speed-type PID controller. Steps S1 to S8 correspond to steps S1 to S8 in FIG.2, respectively. In this compensation, likewise the position type, the disturbance of the internal state of the controller is resolved by replacing the cause of the difference between a difference of the MV calculated by the controller and a difference applied to the MV that is provided actually to the controlled object plant 101 with the cause of the SV difference, and thus a deterioration of the response is prevented.

[0043] In such an event that not the difference applied to the MV that is provided actually to the controlled object plant 101 but the MV that is provided actually into the controlled object plant 101 is obtained, the difference of MV may be calculated from this MV.

[0044] Here, in the case of the position type, the deviation and the integration are contained in Eq.(1), and therefore both the deviation and the integration are compensated. In the case of the speed type, the integration is not contained in Eq.(3), and therefore only the deviation is compensated. In the case of the speed type, it may be considered that, since the actuator 103 calculates the MV from the deviation of MV, the actuator 103 has substantially the integrated value.

[0045] The actuator 103 updates the state by using the difference, which is applied to the MV provided actually to the controlled object plant, in each of the samplings irrespective of the packet loss. No disorder is produced in the actuator 103.

[0046] Next, procedures for setting an MV complementary value in an MV complementing unit 104 will be explained hereunder. FIG.10 is a flowchart showing procedures for setting an MV complementary value in the prior art. In step S1, the controller 201 calculates a manipulated variable U (= MV) in respective samplings. In step S2, the controller 201 transmits the manipulated variable U to the actuator 103 via the network.

[0047] In step S3, it is determined whether or not the actuator 103 could receive the manipulated variable U. If the actuator 103 could receive the manipulated variable U (Yes in step S3), this actuator 103 changes the manipulated variable provided to the controlled object plant 101 to the manipulated variable U in step S4. In contrast, if the actuator 103 could not receive the manipulated variable U due to the packet loss (No in step S3), this actuator 103 compensates for this manipulated variable by using any value, and provides this variable to the controlled object plant 101 in step S4.

[0048] As the way of compensating the value, there are two approaches. One approach is that the preceding value is held to maintain the manipulated variable provided currently to the controlled object plant 101 as it is (the second preceding value is held if the packet loss also occurs in the preceding sampling), and the other approach is that no value is provided to the controlled object plant 101 (the manipulated variable is set to zero) (step S5).

[0049] According to the approach of holding the preceding value, there is such a possibility that, for example, when the set value is changed, the manipulated variable has a very large value only in that sampling. At this time, when the packet loss occurs in the next sampling, this very large manipulated variable is held to disturb the response.

[0050] According to the approach of setting the manipulated variable to zero, the manipulated variable is suddenly set to zero in such a situation that any manipulated variable is provided to the plant up to now, which also disturbs the response.

[0051] FIG.4 is a flowchart showing procedures for setting an MV complementary value according to the present invention. In step S1, the controller 201 calculates the manipulated variable in respective samplings separately as following two values. One value is a first manipulated variable U1 corresponding to a value that is to be held next time, and the other value is a second manipulated variable U2 corresponding to a value except the value that is to be held next time. The manipulated variable U1 and U2 are given by Eq.(5) and Eq.(6).

[0052] In step S2, the controller 201 transmits the manipulated variables U1 and U2 to the actuator 103 via the network 300. In step S3, it is determined whether or not the actuator 103 could receive the manipulated variables U1 and U2. If the actuator 103 could receive the manipulated variables U1 and U2 (Yes in step S3), this actuator 103 changes the manipulated variables to the manipulated variables U1 and U2 and provides the variables U1 and U2 to the controlled object plant 101 in step S4.

[0053] In contrast, if the actuator 103 could not receive the manipulated variables U1 and U2 due to the packet loss (No in step S3), the actuator 103 provides only the manipulated variable U1 out of the manipulated variables U1 and U2 being received last time (the values U1 and U2 being received before the last time if the packet loss occurs in the last time sampling) into the controlled object plant 101 in step S5.

[0054] In this manner, if an amount of SV change in the differentiating operation is assigned to the variable U2 and an amount of remaining changes is assigned to the variable U1, it is possible to resolve the problem that the slightly larger manipulated variable being output when the set value is changed is held, which is also the problem in the conventional approach. The reason is as follows. That is, such an event depends on an amount of SV change in the differentiating operation that, when the set value is changed, the slightly larger manipulated variable is output. Therefore, even when the packet loss occurs in the next sampling, an amount of SV change in the differentiating, operation (the variable U2), which leads to the manipulated variable being increased when the set value is changed, is never provided to the controlled object plant 101.

[0055] FIG.5 is a functional block diagram showing another embodiment of a network control system according to the present invention. The configuration of FIG.5 is different from that of FIG. 1 in that a transmission/reception success/failure detecting unit 103b is added to the actuator 103.

[0056] The actuator 103 may be configured such that the actuator does not transmit the MV trend data having (m+1) pieces of information in each sampling but may transmit the data consisting of only the value that is compensated by the MV completer when the actuator 103 could not receive the MV due to the packet loss. As a result, an amount of information transmitted from the actuator 103 to the controller 201 can be decreased, and thus the communication load can be lessened.

[0057]  In this case, the MV data that the actuator 103 holds in the MV buffering unit 103a only consists of the MV complementary value that is compensated by the MV complementing unit 104 when the actuator could not receive the MV. Hence, the actuator 103 stores the MV complementary value in the MV buffering unit 103a only when this actuator could not receive the MV.

[0058] Also, the actuator 103 transmits the MV complementary value in the MV buffering unit 103a to the controller 201 as the MV trend data answerback. In this case, the transmission/reception success/failure detecting unit 103 detects whether or not this transmission/reception succeeded, and the actuator 103 empties out the MV buffering unit 103a only when the actuator 103 can check that the MV data was transmitted without fail.

[0059] This is because the value becomes unnecessary once such value has reached the controller 201. According to this approach, the processes taken on the accumulator side are increased rather than the case where the actuator 103 transmits simply (m+1) pieces of the data. Therefore, the processing function is required to some extent of the actuator 103.

[0060] Further, the controller 201 can be provided with a function that stops temporarily the transmission of the MV to the actuator 103 until the controller 201 acquires the answer-back of the trend data or the complementary value from the actuator 103.

[0061] FIG.6 is a functional block diagram showing still another embodiment of a network control system according to the present invention. The configuration of FIG.6 is different from that of FIG.1 in that a transmission/reception success/failure detecting unit 102b is added.

[0062] The sensor 102 may be configured such that the sensor does not transmit the PV trend data having (m+1) pieces of information in each of the samplings but may transmit the data consisting of only the latest PV and the PV that has not reached the controller before. As a result, an amount of information transmitted from the sensor 102 to the controller 201 can be decreased, and thus the communication load can be lessened.

[0063] In this case, the PV data that is held by the PV buffering unit 102a of the sensor 102 consists of only the latest PV and the PV that cannot be transmitted to the controller before. The sensor 102 transmits the data in the PV buffering unit 102a to the controller 201. In this event, the transmission/ reception success/failure detecting unit 102b detects whether or not the transmission/reception is performed successfully, and the sensor 102 empties out the PV buffering unit 102a only when the sensor 102 checks that this sensor transmits the data without fail.

[0064] This is because the value is not needed once the value reaches the controller 201. According to this approach, the processes performed on the accumulator side are increased rather than the case where the sensor 102 transmits simply the (m+1) pieces of the data. Therefore, the processing function is required to some extent of the sensor 102.

[0065]  FIG.7 is a functional block diagram showing yet another embodiment of a network control system according to the present invention. This embodiment is characterized in that the controller 201 is integrated with the sensor 102 and only the information transmission between the controller 201 and the actuator 103 is performed via the network 300.

[0066] In this case, the packet loss of the PV never occurs. Therefore, the PV complementing unit, the PV buffering unit in the sensor 102, and the PV buffering unit in the controller 201 are not needed. Also, the information transmitted from the sensor 102 to the controller 201 is not the PV trend data but the latest PV only.

[0067] In order to obtain the PV complementary value from the PV complementing unit 202, the approach of calculating the value predicted based on the transition of PV trend data and then setting the PV complementary value is considered to be common. In this case, when a process model 202a is provided, the PV complementing unit 202 can correct the PV complementary value by predicting the PV from the MV trend data.

Claims

1. A network control system for packet-transmitting a manipulated variable of a controller (201), which is calculated based on a deviation between a processed variable from a sensor (102) that measures a physical quantity of a controlled object plant (101) and a set value at a predetermined sampling period, to an actuator (103) provided on the controlled object plant side via a network (300), and causing the actuator (103) to provide the manipulated variable to the controlled object plant (101), the network control system comprising:

an MV complementing unit (104) adapted to provide a complementary value of the manipulated variable to the actuator (103) when an error occurs in a packet transmission;

characterized by

a first MV buffering unit (103a) adapted to answer back trend data of the manipulated variable being provided to the controlled object plant (101) from the actuator or the complementary value being complemented by the MV complementing unit for the actuator, to the controller (201) via the network (300); and

an MV compensating unit (201c) adapted to correct a calculation of the manipulated variable of the controller (201) based on the trend data or the complementary value being answered back.

2. The network control system according to claim 1, wherein the first MV buffering unit (103a) is provided in the actuator (103), and is adapted to hold only the complementary value of the manipulated variable provided to the actuator (103) by the MV complementing unit (104) when the actuator (103) could not receive the manipulated value due to an error in the packet transmission of the manipulated variable,
the actuator (103) is adapted to transmit only the complementary value of the manipulated variable held in the first MV buffering unit (103a) to the controller (201) via the network (300), and
the MV compensating unit (201 c) is adapted to correct the calculation of the manipulated variable of the controller (201) based on the complementary value of the manipulated variable transmitted by the actuator (103), when an error occurred in the packet transmission of the manipulated variable.

3. The network control system according to claim 1 or 2, further comprises a second MV buffering unit (201 a) provided in the controller (201),
wherein the second MV buffering unit (201a) is adapted to hold (m+1) samples of the manipulated variable that the controller (201) calculated in the current sampling and m samplings prior to the current sampling, and
the controller (201) is adapted to compare the (m+1) samples of the manipulated variable transmitted by the actuator (103) with the (m+1) samples of the manipulated variable held in the second MV buffering unit (201 a), so as to determine whether an error occurred in the packet transmission.

4. The network control system according to any of claims 1 to 3, wherein, when a sample of the manipulated variable calculated by the controller (201) is different from a sample of the manipulated variable actually provided to the controlled object plant (101), the controller (201) is adapted to determine k_max, which is the number of the sampling prior to the current sampling n in which the manipulated variable held by the controller (201) firstly differs from the manipulated variable actually provided into the controlled object plant (101),
to calculate for each of the k_max samplings prior to the current sampling n a virtual set value r_n-k, k = 1, 2, ... k_max, by repeatedly substituting U_n-k of the equation


with the manipulated variable that is actually provided into the controlled object plant (101) in the (n-k)-th sampling, wherein Δt is the sampling period, X_n-k is the processed variable from the sensor in the (n-k)-th sampling, e_n-k-1 is the deviation in the (n-k-1)-th sampling, K_p is the control gain, T, is the integration time and T_D is the differentiation time,
to correct the deviation in the (n-k)-th sampling and the integrated value of the deviation up to the (n-k)-th sampling for each of the k_max samplings prior to the current sampling n using the virtual set value r_n-k of the (n-k)-th sampling, k = 1, 2, ... k_max, and
to update the samples of manipulated variable held by the controller (201) with the samples of the manipulated variable actually provided into the controlled object plant (101).

5. The network control system according to any of claims 1 to 4, wherein the controller (201) is adapted to stop temporarily the transmission of the manipulated variable to the actuator (103) until the controller (201) acquires an answer-back of the trend data or the complementary value.

6. The network control system according to any of claims 1 to 5, wherein the controller (201) is adapted to transmit separately a first manipulated variable corresponding to an amount of change in the set value in a differentiating operation and a second manipulated variable corresponding to an amount of change in the set value except the amount of change in the set value in the differentiating operation, and
the MV complementing unit (104) is adapted to provide a sum of the first manipulated variable and the second manipulated variable to the actuator (103) as the complementary value when the packet transmission is in a normal state, and to provide only the second manipulated variable to the actuator (103) as the complementary value when the packet transmission is in an error state.

7. The network control system according to claim 1, wherein the processed variable measured by the sensor (102) is packet-transmitted to the controller (201) via the network (300), and the network control system further comprises:

a PV complementing unit (202) adapted to provide a complementary value of the processed variable to the controller (201) when an error occurs in the packet transmission; and

a PV buffering unit (102a) adapted to transmit trend data of the processed variable measured by the sensor to the controller (201) via the network (300).

8. The network control system according to claim 7, wherein the PV complementing unit (202) is adapted to correct the complementary value of the processed variable based on the trend data of the manipulated variable.

Ansprüche

1. Netzwerksteuersystem für die Paketübertragung einer manipulierten Variable einer Steuereinrichtung (201), die basierend auf einer Abweichung zwischen einer verarbeiteten Variable von einem Sensor (102), der eine physikalische Größe einer gesteuerten Objektanlage (101) misst, und einem Sollwert bei einer vorbestimmten Abtastungsperiode berechnet wird, an ein Stellglied (103), das auf der Seite der gesteuerten Objektanlage vorgesehen ist, über ein Netzwerk (300), und zum Veranlassen des Stellglieds (103), die manipulierte Variable zu der gesteuerten Objektanlage (101) vorzusehen, wobei das Netzwerksteuersystem umfasst:

eine MV-Komplementierungseinheit (104), die ausgebildet ist zum Vorsehen eines komplementären Werts der manipulierten Variable zu dem Stellglied (103), wenn ein Fehler in einer Paketübertragung auftritt,

gekennzeichnet durch

eine erste MV-Pufferungseinheit (103a), die ausgebildet ist zum Rückantworten von Trenddaten der manipulierten Variable, die zu der gesteuerten Objektanlage (101) von dem Stellglied vorgesehen wird, oder des komplementären Werts, der durch die MV-Komplementierungseinheit für das Stellglied komplementiert wird, an die Steuereinrichtung (201) über das Netzwerk (300), und

eine MV-Kompensierungseinheit (201 c), die ausgebildet ist zum Korrigieren einer Berechnung der manipulierten Variable der Steuereinrichtung (201) basierend auf den Trenddaten oder dem komplementären Wert, die rückgeantwortet werden.

2. Netzwerksteuersystem nach Anspruch 1, wobei die erste MV-Pufferungseinheit (103a) in dem Stellglied (103) vorgesehen ist und ausgebildet ist zum Speichern nur des komplementären Werts der manipulierten Variable, die zu dem Stellglied (103) durch die MV-Komplementierungseinheit (104) vorgesehen wird, wenn das Stellglied (103) die manipulierte Variable aufgrund eines Fehlers in der Paketübertragung der manipulierten Variable nicht empfangen konnte,
wobei das Stellglied (103) ausgebildet ist zum Senden nur des komplementären Werts der manipulierten Variable, der in der ersten MV-Pufferungseinheit (103a) gespeichert ist, an die Steuereinrichtung (201) über das Netzwerk (300), und
die MV-Kompensierungseinheit (201c) ausgebildet ist zum Korrigieren der Berechnung der manipulierten Variable der Steuereinrichtung (201) basierend auf dem komplementären Wert der manipulierten Variable, der durch das Stellglied (103) gesendet wird, wenn ein Fehler in der Paketübertragung der manipulierten Variable aufgetreten ist.

3. Netzwerksteuersystem nach Anspruch 1 oder 2, das weiterhin eine zweite MV-Pufferungseinheit (201 a), die in der Steuereinrichtung (201) vorgesehen ist, umfasst,
wobei die zweite MV-Pufferungseinheit (201a) ausgebildet ist zum Speichern von m+1 Abtastungen der manipulierten Variable, die die Steuereinrichtung (201) in der aktuellen Abtastung berechnet hat, und von m Abtastungen vor der aktuellen Abtastung, und
wobei die Steuereinrichtung (201) ausgebildet ist zum Vergleichen der m+1 Abtastungen der manipulierten Variable, die durch das Stellglied (103) gesendet wird, mit den m+1 Abtastungen der manipulierten Variable, die in der zweiten MV-Pufferungseinheit (201 a) gespeichert sind, um zu bestimmen, ob ein Fehler in der Paketübertragung aufgetreten ist.

4. Netzwerksteuersystem nach einem der Ansprüche 1 bis 3, wobei, wenn eine Abtastung der manipulierten Variable, die durch die Steuereinrichtung (201) berechnet wird, verschieden ist von einer Abtastung der manipulierten Variable, die tatsächlich zu der gesteuerten Objektanlage (101) vorgesehen wird, die Steuereinrichtung (201) ausgebildet ist zum Bestimmen von k_max, das die Zahl der Abtastung vor der aktuellen Abtastung n ist, in welcher sich die durch die Steuereinrichtung (201) gespeicherte manipulierte Variable zuerst von der tatsächlich zu der gesteuerten Objektanlage (101) vorgesehenen manipulierten Variable unterscheidet,
zum Berechnen, für jede der k_max Abtastungen vor der aktuellen Abtastung n, eines virtuellen Sollwerts r_n-k, k = 1, 2, ... k_max durch das wiederholte Ersetzen von U_n-k der Gleichung

durch die manipulierte Variable, die tatsächlich zu der gesteuerten Objektanlage (101) in der (n-k)-ten Abtastung vorgesehen wird, wobei Δt die Abtastungsperiode ist, X_n-k die verarbeitete Variable aus dem Sensor in der (n-k)-ten Abtastung ist, e_n-k-1 die Abweichung in der (n-k-1)-ten Abtastung ist, K_p die Steuerverstärkung ist, T_I die Integrationszeit ist und T_D D die Differentiationszeit ist,
zum Korrigieren der Abweichung in der (n-k)-ten Abtastung und des integrierten Werts der Abweichung bis zu der (n-k)-ten Abtastung für jede der k_max Abtastungen vor der aktuellen Abtastung n unter Verwendung des virtuellen Sollwerts r_n-k der (n-k)-ten Abtastung, k = 1, 2, ... k_max, und
zum Aktualisieren der Abtastungen der durch die Steuereinrichtung (201) gespeicherten Variable mit den Abtastungen der tatsächlich zu der gesteuerten Objektanlage (101) vorgesehenen manipulierten Variable.

5. Netzwerksteuersystem nach einem der Ansprüche 1 bis 4, wobei die Steuereinrichtung (201) ausgebildet ist zum vorübergehenden Stoppen der Übertragung der manipulierten Variable an das Stellglied (103), bis die Steuereinrichtung (201) eine Rückantwort der Trenddaten oder des komplementären Werts erhält.

6. Netzwerksteuersystem nach einem der Ansprüche 1 bis 5, wobei die Steuereinrichtung (201) ausgebildet ist zum separaten Übertragen einer ersten manipulierten Variable in Entsprechung zu einer Änderungsgröße in dem Sollwert in einer Differentiationsoperation und einer zweiten manipulierten Variable in Entsprechung zu einer Änderungsgröße in dem Sollwert mit Ausnahme der Änderungsgröße in dem Sollwert in der Differentiationsoperation, und
die MV-Komplementierungseinheit (104) ausgebildet ist zum Vorsehen einer Summe aus der ersten manipulierten Variable und der zweiten manipulierten Variable zu dem Stellglied (103) als des komplementären Werts, wenn sich die Paketübertragung in einem normalen Zustand befindet, und zum Vorsehen nur der zweiten manipulierten Variable zu dem Stellglied (103) als des komplementären Werts, wenn sich die Paketübertragung in einem Fehlerzustand befindet.

7. Netzwerksteuersystem nach Anspruch 1, wobei die durch den Sensor (102) gemessene verarbeitete Variable an die Steuereinrichtung (201) über das Netzwerk (300) paketübertragen wird und das Netzwerksteuersystem weiterhin umfasst:

eine PV-Komplementierungseinheit (202), die ausgebildet ist zum Vorsehen eines komplementären Werts der verarbeiteten Variable zu der Steuereinrichtung (201), wenn ein Fehler in der Paketübertragung auftritt, und

eine PV-Pufferungseinheit (102a), die ausgebildet ist zum Übertragen von Trenddaten der durch den Sensor gemessenen verarbeiteten Variable an die Steuereinrichtung (201) über das Netzwerk (300).

8. Netzwerksteuersystem nach Anspruch 7, wobei die PV-Komplementierungseinheit (202) ausgebildet ist zum Korrigieren des komplementären Werts der verarbeiteten Variable basierend auf den Trenddaten der manipulierten Variable.

Revendications

1. Système de commande de réseau destiné à transmettre en mode paquet une variable commandée d'un contrôleur (201), qui est calculée sur la base d'un écart entre une variable traitée provenant d'un capteur (102) qui mesure une quantité physique d'une installation d'objet commandé (101) et une valeur de consigne à une période d'échantillonnage prédéterminée, à un module d'actionnement (103) fourni sur le côté de l'installation d'objet commandé, par l'intermédiaire d'un réseau (300), et amenant le module d'actionnement (103) à fournir la variable commandée à l'installation d'objet commandé (101), le système de commande de réseau comprenant :

une unité de complémentation de variable MV (104) apte à fournir une valeur complémentaire de la variable commandée au module d'actionnement (103) lorsqu'une erreur se produit dans une transmission de paquets ;

caractérisé par

une première unité de mise en mémoire tampon de variable MV (103a) apte à retransmettre des données de tendance de la variable commandée fournie à l'installation d'objet commandé (101) à partir du module d'actionnement, ou la valeur complémentaire complétée par l'unité de complémentation de variable MV pour le module d'actionnement, au contrôleur (201), par l'intermédiaire du réseau (300) ; et

une unité de compensation de variable MV (201c) apte à corriger un calcul de la variable commandée du contrôleur (201) sur la base des données de tendance retransmises ou de la valeur complémentaire retransmise.

2. Système de commande de réseau selon la revendication 1, dans lequel la première unité de mise en mémoire tampon de variable MV (103a) est fournie dans le module d'actionnement (103) et est apte à ne conserver que la valeur complémentaire de la variable commandée fournie au module d'actionnement (103) par l'unité de complémentation de variable MV (104), lorsque le module d'actionnement (103) n'a pas pu recevoir la valeur commandée sous l'effet d'une erreur dans la transmission en mode paquet de la variable commandée ;
le module d'actionnement (103) est apte à ne transmettre que la valeur complémentaire de la variable commandée conservée dans la première unité de mise en mémoire tampon de variable MV (103a), au contrôleur (201), par l'intermédiaire du réseau (300) ; et
l'unité de compensation de variable MV (201c) est apte à corriger le calcul de la variable commandée du contrôleur (201) sur la base de la valeur complémentaire de la variable commandée transmise par le module d'actionnement (103), lorsqu'une erreur est survenue dans la transmission en mode paquet de la variable commandée.

3. Système de commande de réseau selon la revendication 1 ou 2, comprenant en outre une seconde unité de mise en mémoire tampon de variable MV (201a) fournie dans le contrôleur (201) ;
dans lequel la seconde unité de mise en mémoire tampon de variable MV (201a) est apte à conserver (m + 1) échantillons de la variable commandée que le contrôleur (201) a calculé dans l'échantillonnage en cours et m échantillonnages avant l'échantillonnage en cours ; et
le contrôleur (201) est apte à comparer les (m + 1) échantillons de la variable commandée transmise par le module d'actionnement (103) aux (m + 1) échantillons de la variable commandée conservée dans la seconde unité de mise en mémoire tampon de variable MV (201a), de manière à déterminer si une erreur est survenue dans la transmission en mode paquet.

4. Système de commande de réseau selon l'une quelconque des revendications 1 à 3, dans lequel, lorsqu'un échantillon de la variable commandée calculée par le contrôleur (201) est différent d'un échantillon de la variable commandée effectivement fournie à l'installation d'objet commandé (101), le contrôleur (201) est apte à déterminer une valeur k_max qui correspond au numéro de l'échantillonnage avant l'échantillonnage en cours n dans lequel la variable commandée conservée par le contrôleur (201) diffère en premier lieu de la variable commandée fournie dans l'installation d'objet commandé (101) ; et
à calculer, pour chacun des k_max échantillonnages avant l'échantillonnage en cours n, une valeur de consigne virtuelle r_n-k, k = 1, 2, ..., K_max, en substituant, de manière répétée, U_n-k de l'équation ci-dessous

par la variable commandée qui est effectivement fournie dans l'installation d'objet commandé (101) dans le (n - k)-ième échantillonnage, dans lequel Δt correspond à la période d'échantillonnage, X_n-K correspond à la variable traitée en provenance du capteur dans le (n - k)-ième échantillonnage, e_n-K-1 correspond à l'écart dans le (n - k - 1)-ième échantillonnage, K_p correspond au gain de commande, T_I correspond au temps d'intégration et T_D représente le temps de différenciation ;
à corriger l'écart dans le (n - k)-ième échantillonnage et la valeur intégrée de l'écart jusqu'au (n - k)-ième échantillonnage pour chacun des k_max échantillonnages avant l'échantillonnage en cours n, en utilisant la valeur de consigne virtuelle r_n-k du (n - k)-ième échantillonnage, k = 1, 2, ..., k_max ; et
à mettre à jour les échantillons de la variable commandée conservée par le contrôleur (201) avec les échantillons de la variable commandée effectivement fournie dans l'installation d'objet commandé (101).

5. Système de commande de réseau selon l'une quelconque des revendications 1 à 4, dans lequel le contrôleur (201) est apte à interrompre temporairement la transmission de la variable commandée au module d'actionnement (103) jusqu'à ce que le contrôleur (201) acquiert une retransmission des données de tendance ou de la valeur complémentaire.

6. Système de commande de réseau selon l'une quelconque des revendications 1 à 5, dans lequel le contrôleur (201) est apte à transmettre séparément une première variable commandée correspondant à une quantité de changement dans la valeur de consigne au cours d'une opération de différenciation, et une seconde variable commandée correspondant à une quantité de changement dans la valeur de consigne hormis la quantité de changement dans la valeur de consigne au cours de l'opération de différenciation ; et
l'unité de complémentation de variable MV (104) est apte à fournir une somme de la première variable commandée et de la seconde variable commandée au module d'actionnement (103) en tant que valeur complémentaire lorsque la transmission en mode paquet est dans un état normal, et à fournir la seconde variable commandée au module d'actionnement (103) en tant que valeur complémentaire lorsque la transmission en mode paquet est dans un état d'erreur.

7. Système de commande de réseau selon la revendication 1, dans lequel la variable traitée mesurée par le capteur (102) est transmise en mode paquet au contrôleur (201) par l'intermédiaire du réseau (300), et le système de commande de réseau comprend en outre :

une unité de complémentation de valeur PV (202) apte à fournir une valeur complémentaire de la variable traitée au contrôleur (201) lorsqu'une erreur se produit dans la transmission en mode paquet ; et

une unité de mise en mémoire tampon de valeur PV (102a) apte à transmettre des données de tendance de la variable traitée mesurée par le capteur, au contrôleur (201), par l'intermédiaire du réseau (300).

8. Système de commande de réseau selon la revendication 7, dans lequel l'unité de complémentation de valeur PV (202) est apte à corriger la valeur complémentaire de la variable traitée sur la base des données de tendance de la variable commandée.

Drawing